A Hybrid approach for milk collection using trucks and trailers

2014 
(ProQuest: ... denotes formulae omitted.)1. IntroductionCollection of milk from farms is a challenging vehicle routing problem. Milk is perishable so it must be collected from the farms and transported to the production plants as soon as possible. This problem is handled differently from country to country due to the varying nature of geographical and environmental conditions.Geographically, milk farms are mostly located in the rural areas of a country. Due to the perishable nature of milk and the limited storage capacities at farms, it must be collected and transported to production factories within a given time limit. The production quantity of milk depends on the size of each farm, as the number of cows varies.Trucks and trailers with different capacities/sizes are used for collection and are owned or hired by the company. Transportation companies usually desire to acquire fleet of vehicles which are not too expensive but have a high degree of reliability and durability and that will be optimally utilized. It is not economical to invest in a fleet mix which remains idle or unutilized in a given period. Thus the objective of the company is to choose an appropriate combination of trucks and trailers which could reduce the total transportation cost. In addition to the fleet composition problem, there are also weights and accessibility restrictions on the roads that make this problem more difficult. As a result of this added complexity in the problem constraints, the distribution planning decisions are harder to make. Transportation cost is a central aspect to be considered in distribution planning. It generally involves driver's cost, truck and trailer's acquisition cost, variable routing cost, toll road/ferry taxes and others. Thus, milk collection normally incurs a major transportation cost. Still, this type of problem is solved manually in most countries, like many other problems. In this paper, we provide a simple cost effective technique to this vehicle routing planning and distribution problem.This paper describes a real world problem where TINE SA, the leading dairy company in Norway, collects and delivers raw milk from different farms to the production factories. The methodology developed and explained in this paper provides an economic solution to this type of problem. The special case of Truck and Trailer Vehicle Routing Problem considers four major aspects: fleet composition, multiple depot, routing decisions and location of parking places for trailers. Due to road limitations, a loaded truck with a trailer cannot be driven on roads leading to the farms and consequently the routes are constructed such that the trailer must be temporarily parked in designated parking places while the truck has to be driven on a sub route to collect milk at the farms.Routing decisions are concerned with assigning farms to a truck that is able to fulfill the farm's demand while satisfying the road restriction constraints. Farms may be serviced by one of three possible depots in this specific region. Here, depot is represented as a dairy/plant. Various types of trucks with or without trailers are available on the production factory. The capacity of each truck varies in terms of volume and weight.The remainder of this paper is structured as follows. In Section 2, a brief overview of research related to the milk collection problem is presented. Section 3 gives a detailed description of the real world problem while the solution methodology is explained in Section 4. Information about real world data and tuning of parameters are demonstrated in section 5. Computational results are presented in Section 6, and conclusions and future recommendations have been made in Section 7.2. Literature ReviewThe real world problem explained in this paper is a special type of truck and trailer vehicle routing problem which has not been directly addressed in the research literature. In the last two decades, only a few research works with some indirect relationship with the problem have been published. …
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    1
    Citations
    NaN
    KQI
    []